Rectal cancer preoperative concurrent neoadjuvant radiotherapy and chemotherapy effect evaluation system and method based on big data analysis MRI images

Abstract

The invention relates to a rectal cancer preoperative concurrent neoadjuvant radiotherapy and chemotherapy effect evaluation system and method. The evaluation system comprises: an image acquisition unit for acquiring MRI images before neoadjuvant radiotherapy and chemotherapy for patients with the locally advanced rectal cancer, and dividing the rectal cancer patients into a training set, a checkset and a test sets as input image data; an image annotation unit for performing data annotation on MRI images of the training set, the check set and the test set; a convolutional neural network construction unit for constructing a first convolutional neural network model; and a convolutional neural network model training unit for acquiring a second convolutional neural network model for evaluating the rectal cancer preoperative concurrent neoadjuvant radiotherapy and chemotherapy effect. The rectal cancer preoperative concurrent neoadjuvant radiotherapy and chemotherapy effect evaluation system has many advantages such as high accuracy, short time-consuming, long working duration, objective and three-dimensional result.

Description

technical field [0001] The present invention relates to a system and method for evaluating curative effect of preoperative neoadjuvant chemoradiotherapy for rectal cancer, in particular to a system and method for evaluating curative effect of preoperative neoadjuvant chemoradiotherapy for rectal cancer based on big data analysis of MRI images. Background technique [0002] Deep learning is currently the most suitable and widely used algorithm for image recognition and speech analysis in the field of artificial intelligence. Its inspiration comes from the working mechanism of the human brain. It is to automatically extract features from externally input data by establishing a convolutional neural network. , so that the machine can understand the learning data, obtain information and output. At present, artificial intelligence based on deep learning has been applied in various industries, including speech recognition, face recognition, car logo recognition, handwritten Chinese...

AI Technical Summary

Problems solved by technology

At present, doctors are trained clinically through clinical practice and professional knowledge. Qualified doctors manually review pre-treatment and post-treatment MRI images, and combine their long-term accumulated clinical diagnosis experience to make analysis and diagnosis. The accuracy rate is not high, time-consuming, and continuous work. time is limited

This method of reading images with the naked eye is closely related to subjective factors such as the doctor's own experience, working status, and subjective emotions, and is prone to missed diagnosis, misdiagnosis, and controversial cases.

In addition, post-treatment images are more difficult to identify due to factors such as intestinal wall fibrosis, thickening, and CRT-induced inflammatory infiltration after CRT, which is also a major reason for the difficulty in accurately evaluating the efficacy

In addition, it takes a long time to train a qualified doctor, and the accumulation of clinical experience is indispensable, and it is inevitable to be affected by subjective emotions and working conditions during the long-term work process. Many factors may affect Patient treatment plan development

[0006] The existing radiomics evaluation of the efficacy of neoadjuvant therapy for rectal cancer basically focuses on the tumor area, while ignoring the information on the intestinal wall and surrounding areas where the tumor is likely to invade. However, the difficulty in the diagnosis and treatment of tumors lies in its invasiveness. Therefore, using a single image of the tumor area to evaluate the curative effect has limitations in accuracy and reliability.

Images